Design and development of ontology based knowledge management system for software risks

Abstract

A Knowledge Management System is an integrated multifunctional software framework that is intended to assists, via knowledge processing functions, and supports all main knowledge management and knowledge processing activities, such as capturing, organising, classifying, understanding, debugging, editing, finding, retrieving, disseminating, transferring and sharing knowledge. The research work presented in this thesis provides an ontology based knowledge management solution for Software Risk Management (SRM). SRM is a discipline, whose objectives are to identify, address and eliminate software risk items, before they become either threats to successful software operation, or major sources of expensive software rework. As part of this research, ontology for software risk management, named SRMONTO is designed and developed; it represents the knowledge of Software Risk Management. The research work presented in this thesis is based mainly on the SRMONTO. Its main goal is to provide project analysts with a web service, to aid them in identifying and assessing the risks involved in the project, allowing them to plan for avoiding risks, and controlling and monitoring risks. The motivation of this work is to provide a major incentive for the industry by reducing the cost and the time spent on analyzing projects, and removing the unnecessary resources spent on non-feasible projects. The developed SRMONTO was evaluated both by a quantitative analysis and a qualitative analysis. The quantitative analysis is based on class metrics, property metrics, ratio metrics and axioms metrics. The ontology based web service for software risk management, which has been developed as part of this research work was evaluated by two different case studies. Finally a complete evaluation of the FAOMS was carried out. Two different versions each of the sub ontologies of the SRMONTO were taken into account for testing and evaluating the system using the measures Precision, Recall and F-measure. newline newline newline